Redmine

When trying to open one page, there is an error:

2023/04/24 17:52:46 [error] 2222777#2222777: *11115815 upstream prematurely closed connection 
while reading response header from upstream, client: [ip], server: [serverName], request: 
"GET /projects/projectName/repository/repoName/revisions/trunk/show/db/migr HTTP/2.0", 
upstream: "http://127.0.0.1:8084/projects/projectName/repository/repoName/revisions/trunk/show/db/migr", host: "hostName" 

The upstream is Redmine itself, in the log:

E, [2023-04-24T17:52:46.340940 #848197] ERROR -- : worker=1 PID:110293 timeout (61s > 60s), killing
E, [2023-04-24T17:52:46.354092 #848197] ERROR -- : reaped #<Process::Status: pid 110293 SIGKILL (signal 9)> worker=1
I, [2023-04-24T17:52:46.451416 #117117]  INFO -- : worker=1 ready

I thought it was because of some commit that no longer exists, so I disabled and reconnected the necessary repository - it didn't help. I executed:

rake redmine:fetch_changesets RAILS_ENV=production

- it didn't help.
I tried to change the unicorn timeout from 60 to 120 seconds, and I didn't understand how to do it except for a reboot. I killed the unicorn process and restarted it - unicorn -c config/unicorn.rb -E production -D - but Redmine stopped working.
I put the timeout back in place and restarted the unicorn again - it didn't help. I rebooted the server - Redmine works. How to properly reload unicorn settings? Or could it be something else?